On Thu, 14 Mar 2002, Peter Jay Salzman wrote:
> here's how these numbers are generated
>
> int trials = some big integer;
>
> while (i < trials)
> {
> energy per site += some calculation;
> }
> energy per site /= trials*othernumbers
>
> i hope this makes it clear that it can't be a rounding error. energy
> per site accumulates some large value which gets divided by a common
> factor at the end of the calculation.
It's hard to tell without knowing the dataset (or at least their
approximate range). If, for example, "energy_per_site" becomes a large
number relative to "some_calculation" during some part of the loop
iteration, then rounding errors can occur.
I'm curious to see if there can indeed be a non-rounding error explanation
to all this. Could you send me the complete program with dataset for the
above example? Thanks!
Mark
--
Mark K. Kim
http://www.cbreak.org/
PGP key available upon request.
_______________________________________________
vox-tech mailing list
[EMAIL PROTECTED]
http://lists.lugod.org/mailman/listinfo/vox-tech
